[Federal Register Volume 78, Number 213 (Monday, November 4, 2013)]
[Notices]
[Pages 65963-65970]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2013-26266]


-----------------------------------------------------------------------

DEPARTMENT OF COMMERCE

International Trade Administration


Antidumping Proceedings: Announcement of Change in Department 
Practice for Respondent Selection in Antidumping Duty Proceedings and 
Conditional Review of the Nonmarket Economy Entity in NME Antidumping 
Duty Proceedings

AGENCY: Import Administration, International Trade Administration, 
Department of Commerce.

ACTION: Change in Practice to the Department's Respondent Selection in 
Certain Antidumping Duty Proceedings and Elimination of Conditional 
Review of the NME Entity.

-----------------------------------------------------------------------

SUMMARY: The Department of Commerce (``Department'') is hereby refining 
its practice with respect to the methodology for respondent selection 
in certain antidumping (``AD'') proceedings. Specifically, the 
Department is making changes to its current practice in antidumping 
administrative reviews for (1) respondent selection; and (2) 
conditional review of the NME entity. Normally, the Department makes 
these

[[Page 65964]]

types of changes to its practice in the context of its case 
proceedings, on a case-by-case basis.\1\ For these particular changes 
in practice, the Department sought comments in advance of making 
changes in practice. However, the Department expects to continue to 
consider, and make changes in practice, as necessary, in the context of 
its proceedings based upon comments from interested parties submitted 
in the course of such proceedings.\2\
---------------------------------------------------------------------------

    \1\ In the context of its proceedings, Commerce is entitled to 
change its practice and adopt a new administrative practice provided 
it explains the basis for the change, and the change is a reasonable 
interpretation of the statute. Saha Thai Steel Pipe Company v. 
United States, 635 F.3d 1335, 1341 (2011).
    \2\ In particular, under 19 U.S.C. 1677f-1(b), the authority to 
select ``statistically valid samples rests exclusively with the 
administering authority.'' Commerce must retain the ability to alter 
its sampling methodology in each case, as is clear from the above 
provision that Commerce ``shall, to the greatest extent possible, 
consult with the exporters and producers regarding the method to be 
used to select exporters, producers, or types of products under this 
section.''

DATES: Applicability date: The Department expects to apply these 
changes in practice in AD administrative reviews for which the notice 
of opportunity to request an administrative review is published on or 
---------------------------------------------------------------------------
after December 4, 2013.

FOR FURTHER INFORMATION CONTACT: Shauna Biby, International Trade 
Analyst, Import Administration, U.S. Department of Commerce, at 202-
482-4267.

SUPPLEMENTARY INFORMATION: The Department is hereby refining its 
practice with respect to the methodology for respondent selection in 
certain AD proceedings. Specifically, the Department intends to select 
respondents by sampling where certain criteria are met in AD 
administrative reviews. Further, while considering issues related to 
respondent selection and sampling, the Department has also reconsidered 
its practice of ``conditionally'' reviewing the nonmarket economy 
(``NME'') entity. In an administrative review of an AD order, the 
Department's current practice is to consider the NME entity to be 
``conditionally'' under review. This means that even absent a request 
for review of the entity, the entity will become subject to review if 
an exporter subject to the review does not demonstrate that it is 
separate from the entity, and the entity's entries will be potentially 
subject to a new cash deposit and assessment rate. The Department has 
determined to discontinue such conditional reviews. If interested 
parties wish to request a review of the entity, such a request must be 
made in accordance with the Department's regulations.
    The Department notes that in June 2005, it requested and received 
comments on the timing of assessment instructions for AD orders 
involving NME cases.\3\ Many commenters expressed support for a 
practice that would not delay assessment instructions of certain 
entries based on the Department's conditional review of the NME 
entity.\4\ Although the Department did not revise its practice with 
respect to conditional review of the NME entity at that time, the 
Department's experience to date indicates that there is no ongoing 
benefit to be achieved in maintaining conditional review of the entity. 
Furthermore, by eliminating the practice of conditional review, the 
Department eliminates an unnecessary delay in liquidation.
---------------------------------------------------------------------------

    \3\ See Timing of Assessment Instructions for Antidumping Duty 
Orders Involving Non-Market Economy Countries, 70 FR 35634 (June 21, 
2005).
    \4\ See public comments received July 15, 2005, available at 
http://ia.ita.doc.gov/download/nme-assessment/nme-assessment-timing.html.
---------------------------------------------------------------------------

    The notice-and-comment requirements of the Administrative 
Procedures Act do not apply to interpretive rules, general statements 
of policy or procedure, or practice. 5 U.S.C. 553(b)(3)(A). Although 
the notice-and-comment requirements of the Administrative Procedure Act 
do not apply, the Department provided an opportunity for the public to 
comment on the Department's proposed refinement to respondent selection 
in a notice published on December 16, 2010; and for the public to 
comment on the Department's practice with respect to the timing of 
assessment instructions in NME cases in a notice published on June 21, 
2005.

Sampling Methodology

Background

    On December 16, 2010, the Department proposed a refinement to its 
practice regarding its methodology for respondent selection in AD 
proceedings.\5\ As explained in the Proposed Methodology, when the 
number of producers/exporters (``companies'') involved in an AD 
investigation or review is so large that the Department finds it 
impracticable to examine each company individually, the Department has 
the statutory authority to limit its examination to: (1) A sample of 
exporters, producers, or types of products that is statistically valid 
based on the information available to the administering authority at 
the time of selection, or (2) exporters and producers accounting for 
the largest volume of subject merchandise from the exporting country 
that can reasonably be examined.\6\ The Department has, to date, 
generally used the second option in proceedings in which limited 
examination has been necessary. One consequence of this is that 
companies under investigation or review with relatively small import 
volumes have effectively been excluded from individual examination. 
Over time, this creates a potential enforcement concern in AD 
administrative reviews because, as exporters accounting for smaller 
volumes of subject merchandise become aware that they are effectively 
excluded from individual examination by the Department's respondent 
selection methodology, they may decide to lower their prices as they 
recognize that their pricing behavior will not affect the AD rates 
assigned to them. Sampling such companies under section 777A(c)(2)(A) 
of the Tariff Act of 1930, as amended (the ``Act''), is one way to 
address this enforcement concern.
---------------------------------------------------------------------------

    \5\ See Proposed Methodology for Respondent Selection in 
Antidumping Proceedings; Request for Comment, 75 FR 78678 (December 
16, 2010) (``Proposed Methodology'').
    \6\ See sections 777A(c)(2)(A) and (B) of the Act.
---------------------------------------------------------------------------

    The statute requires that the sample be ``statistically valid.'' 
\7\ The Department has interpreted this as referring to the manner in 
which the Department selects respondents.\8\ Therefore, to ensure the 
statistical validity of samples, in the Proposed Methodology, the 
Department proposed employing a sampling technique that: (1) Is random; 
(2) is stratified; and (3) uses probability-proportional-to-size 
(``PPS'') samples. Random selection ensures that every company has a 
chance of being selected as a respondent and captures potential 
variability across the population. Stratification by import volume 
ensures the participation of companies with different ranges of import 
volumes in the review, which is key to addressing the enforcement 
concern identified above. Finally, PPS samples ensure that the 
probability of a company being chosen as a respondent is proportional 
to its share of imports in the respective stratum.
---------------------------------------------------------------------------

    \7\ See section 777A(c)(2)(A) of the Act.
    \8\ See Brake Rotors From the People's Republic of China: Final 
Results and Partial Rescission of the 2004/2005 Administrative 
Review and Notice of Rescission of 2004/2005 New Shipper Review, 71 
FR 66304 (November 14, 2006) and accompanying Issues and Decision 
Memorandum at Comment 1A (``Brake Rotors'').
---------------------------------------------------------------------------

The Department's Sampling Methodology

    In general, the Department will normally rely on sampling for

[[Page 65965]]

respondent selection purposes in AD administrative reviews \9\ when the 
following conditions are met: (1) There is a request by an interested 
party for the use of sampling to select respondents; (2) the Department 
has the resources to examine individually at least three companies for 
the segment; (3) the largest three companies (or more if the Department 
intends to select more than three respondents) by import volume of the 
subject merchandise under review account for normally no more than 50 
percent of total volume; and (4) information obtained by or provided to 
the Department provides a reasonable basis to believe or suspect that 
the average export prices and/or dumping margins for the largest 
exporters differ from such information that would be associated with 
the remaining exporters.\10\
---------------------------------------------------------------------------

    \9\ This sampling methodology has been developed for AD 
administrative reviews, not AD investigations, or countervailing 
duty investigations or reviews.
    \10\ This information may include for example: (1) Company 
margins from previous segments of the proceeding; (2) market and 
company pricing information; (3) the nature and structure of the 
foreign industry in question, including cost structure and/or actual 
pricing data; and (4) the U.S. Customs and Border Protection import 
entry database.
---------------------------------------------------------------------------

Accuracy of the Sampling Method

    Many of the commenters who oppose the proposed methodology focus on 
the issue of accuracy, and query how a small sample can be 
``statistically valid'' within the meaning of the statute. However, in 
a previous proceeding, the Department explained that the phrase 
``statistically valid'' in section 777A(c)(2)(A) of the Act refers to 
the manner or process by which the sample is taken, not the sample 
results.\11\ In that proceeding, the Department explained that ``the 
phrase `statistically valid sample' was added to the statute in 1994 
merely to conform the language of the statute with that of the World 
Trade Organization (``WTO'') AD Agreement (Agreement on Implementation 
of Article VI of the General Agreement on Tariffs and Trade 1994), and 
is not different in substance from the phrase `generally recognized 
sampling techniques' used in the Act prior to the URAA.'' \12\ The 
Department determined that the ``statistical validity'' of the sample 
``refers only to the manner in which the respondents are selected, and 
not to the size of the sample under review.'' \13\
---------------------------------------------------------------------------

    \11\ See Brake Rotors, 77 FR 66304 and accompanying Issues and 
Decision Memorandum at Comment 1A.
    \12\ Id., (citing Statement of Administrative Action, H.R. Rep. 
No. 103-316, at 872 (1994)).
    \13\ Id.
---------------------------------------------------------------------------

Statistical Validity of the Department's Sampling Method

    The statistical tools in the methodology described herein satisfy 
the requirements for statistical validity. The population average 
(mean) dumping margin of concern to the Department is the export trade-
weighted average dumping margin across all firms (exporters under 
review). Because this trade-weighted average margin, in turn, is 
equivalent to the stratum-weighted average of the stratum means, the 
estimation of the population mean equates to estimation of the stratum 
means. Each stratum mean is estimated on the basis of a PPS-based 
sample mean,\14\ which accounts for the variance in trade shares across 
exporters in the stratum and is, therefore, an unbiased estimator of 
the stratum mean in the sense that there is no systematic error 
associated with repeated sampling. Without PPS sampling, the sample 
mean would be over-weighted toward smaller-exporter margins and a bias 
would result. PPS sampling removes this bias.
---------------------------------------------------------------------------

    \14\ The sample mean is the arithmetic average of the data 
values in the sample. For a sample of ten numbers, the sample mean 
is (x1 +x2. . . .)/10. In the AD respondent 
sampling context, the sample mean for a stratum is the simple 
average of the dumping margins of the sampled respondents from the 
stratum.
---------------------------------------------------------------------------

    Finally, stratification of the sample population into appropriate 
size categories, e.g., small, medium and large-sized exporters by 
import volume, ensures a maximum degree of cross-sectional 
representation of the population in the sample.

Definition of Sampling Population

    Currently, the Department generally chooses companies for 
individual examination based on import volumes reported in case-
specific U.S. Customs and Border Protection (``CBP'') import data. It 
also assigns an AD rate to all other companies that are not selected 
for individual examination. The Department currently does not require 
any evidence of shipment from a non-selected company before making its 
respondent-selection decision. However, in the sampling context, the 
existence of shipments will be required in order to both define the 
population, and if the company is selected, establish a dumping margin 
for the company. Therefore, the Department will normally use CBP data 
as the basis for the volume of subject merchandise and expects to 
define the population from which to sample as all companies named in a 
review with shipments of subject merchandise.
    In NME cases, only those exporters who receive a separate rate will 
be included in the sample population. Companies that do not receive a 
separate rate will not be subject to review pursuant to the elimination 
of the conditional review of the NME entity practice described below. 
Therefore, in order to establish the appropriate sample population at 
the time of the sampling selection, it is necessary for the Department 
to make its determinations regarding the separate rate status of the 
companies under review before the sample is determined. For the purpose 
of constructing the sample rate, the Department expects that companies' 
separate rate status will remain unchanged once the sample is 
determined.

Calculating and Assigning Sample Rates

    After examination of selected respondents by the sampling method, 
the Department will need to assign a rate to all non-selected 
companies. To do so, the Department will calculate a ``sample rate,'' 
based upon an average of the rates for the selected respondents, 
weighted by the import share of their corresponding strata. The 
respondents selected for individual examination through the sampling 
process will receive their own rates; all companies in the sample 
population who were not selected for individual examination will 
receive the sample rate.

Implementation of Sampling Methodology

    The Department expects to implement the sampling methodology in the 
context of its administrative reviews by providing interested parties 
with notice of the schedule for submissions related to sampling on a 
case-by-case basis. The Department is publishing concurrently with this 
notice a proposed rule to amend section 351.301 of its regulations, 
``Time limits for submission of factual information,'' to implement 
procedural changes, as needed, with respect to submissions related to 
sampling in antidumping administrative reviews.
    In sum, the rule proposes to require interested parties to submit 
requests for the Department to conduct sampling in antidumping duty 
administrative reviews together with their comments on CBP data within 
seven days following the release of the CBP data, unless otherwise 
specified. The rule proposes that the submission include: (1) A request 
that the Department conduct sampling; and (2) factual

[[Page 65966]]

information \15\ and comments on whether this factual information 
provides a reasonable basis to believe or suspect that the average 
export prices and/or dumping margins for the largest exporters differ 
from such information that would be associated with the remaining 
exporters. Under the proposed rule, if an interested party were to 
submit a request for the Department to conduct sampling, all other 
interested parties will then have a ten-day comment period and a five-
day rebuttal period to comment on the sampling request.\16\
---------------------------------------------------------------------------

    \15\ A detailed description of what this information may include 
is listed in footnote 10 under ``The Department's Sampling 
Methodology'' section of this Federal Register notice.
    \16\ In NME cases, parties must submit their separate rate 
applications or certifications no later than 60 days after the 
notices of initiation of the reviews are published, unless otherwise 
specified in the notices of initiation.
---------------------------------------------------------------------------

    Apart from the proposed rule, in cases in which the Department 
determines to sample for respondent selection, it expects to conduct 
the sampling following the conclusion of the 90-day period for 
withdrawal of requests for administrative reviews under 19 CFR 
351.213(d)(1). In cases in which the Department decides to sample, the 
Department does not expect to exercise its discretion to extend the 90-
day period for withdrawal of review requests.

Comments and Responses

    The Department received 18 comments on the proposed use of sampling 
for selecting mandatory respondents. A summary of these comments are 
presented below and have been grouped by the issues raised in the 
submissions. The Department's response follows immediately after each 
comment.

Issue: Statutory and International Requirements, Including That of 
``Statistical Validity''

    Some commenters generally support the increased use of sampling, 
with several commenters noting that the proposed methodology is 
consistent with statutory requirements. Citing the Statement of 
Administrative Action (``SAA'') and previous instances in which the 
Department has sampled, several commenters note that the Department is 
only required to use a methodology ``designed to give representative 
results based on the facts known at the time of sampling.'' Further, 
the Department must contend with limited time and resources and has the 
discretion under the law to devise an appropriate sampling methodology. 
Other commenters note that the Department should retain as much 
flexibility as possible, and should not confine itself to one sampling 
methodology for all cases and industries.
    Other commenters raised a number of concerns with whether the 
proposed methodology meets the Department's statutory and international 
obligations. Further, these commenters generally questioned whether the 
proposed methodology is ``statistically valid,'' arguing that the 
Department must make some finding about the degree of precision it will 
require. Specifically, there is no reference to size or ``precision'' 
of the sample in the proposed methodology. Some commenters asserted 
that a ``statistically valid sample'' is a higher standard than a 
``generally recognized sampling technique.'' Moreover, ``statistically 
valid'' must ``include the key ideas of the size of the sample and the 
relationship of the sample to the whole.'' The core problem, some 
commenters noted, is that, in most cases, the Department does not have 
the resources to investigate the large number of companies that would 
be required to make the sample statistically valid. These commenters 
generally note that sample size cannot be fixed at the start, but 
rather one determines sample size based on three factors: the number of 
companies whose behavior is being measured, the margin of error likely 
to result, and finally, the ``confidence'' level desired.\17\ These 
commenters assert that 90 or 95 percent is a typical confidence level. 
In sum, sample size must be large enough to permit a statistically 
valid inference. The statute therefore provides an alternative: Choose 
the largest exporters. This method, the commenters assert, will 
normally yield the most accurate and comprehensive results.
---------------------------------------------------------------------------

    \17\ ``Confidence level'' relates to the probability that a 
sample-based estimate falls within specified error limits of the 
estimated parameter value, and the range of values defined by an 
estimate plus or minus the specified error limit is a ``confidence 
interval.''
---------------------------------------------------------------------------

    With respect to the Department's international obligations, one 
commenter submitted that any respondent selection practice must comply 
with the Antidumping Agreement (``ADA'') Article 9.3, under which a 
company's margin is linked to its behavior, stating further that the 
proposed sampling methodology lacks any such link. Further, the 
selection process must not produce results that deprive respondents of 
the right to revocation under Articles 11.1 and 11.3 of the ADA. 
Companies not selected as mandatory respondent have no opportunity to 
assert these rights.
    The Department's response: The Department addresses the majority of 
these issues herein and otherwise will address any particular 
circumstances as they arise on a case-by-case basis. Specifically, the 
statute requires that the sample be ``statistically valid.'' The 
Department has interpreted this as referring to the manner in which the 
Department selects respondents and not to the size of the sample or 
precision of the sample results. Therefore, to ensure the statistical 
validity of samples, the Department will employ a sampling technique 
that: (1) Is random; (2) is stratified; and (3) uses PPS samples. 
Random selection ensures that every company has a chance of being 
selected as a respondent and captures potential variability across the 
population. Stratification by import volume ensures the participation 
of companies with different ranges of import volumes in the review, 
which is key to addressing the enforcement concerns identified herein. 
Finally, PPS samples ensure that the probability of a company being 
chosen as a respondent is proportional to its share of imports in the 
respective stratum. The Department intends to address any further 
comments on the statistical validity of its sampling methodology on a 
case-by-case basis as they arise. Finally, the Department will address 
any specific concerns with respect to revocation as they arise on a 
case-by-case basis.

Issue: Clarifying the Rationale for Increased Use of Sampling

    Several commenters asserted that the Department failed to define 
the objective of its sampling proposal nor had it described or 
explained what benefits it perceives from sampling, for example, how 
sampling would advance any statutory or policy objective. Noting 
resource constraints, one commenter urged the Department to recall its 
authority under the Act to simplify and streamline procedures, 
including the use of averaging and statistically valid samples. 
Further, these commenters generally asserted that the Department should 
maintain its preference for selecting the largest exporters based on 
volume, which will result in ``dumping margins that more accurately 
reflect the pricing of subject merchandise in the U.S.''
    The Department's response: As noted herein, the Department has, to 
date, generally chosen the largest respondents in proceedings in which 
limited examination has been necessary. One consequence of this is that 
companies under review with relatively small

[[Page 65967]]

import volumes have generally been effectively excluded from individual 
examination. This creates a potential enforcement concern in AD 
administrative reviews because, as exporters accounting for smaller 
volumes of subject merchandise become aware that they are effectively 
excluded from individual examination by the Department's respondent 
selection methodology, they may decide to lower their prices as they 
recognize that their pricing behavior will not impact the AD rates 
assigned to them. Sampling companies under section 777A(c)(2)(A) of the 
Act is one way to remedy this enforcement concern. Therefore, the 
Department is exercising its discretion to use sampling in its 
respondent selection procedures.

Issue: The Use of CBP Data and Other Issues Regarding Import Shares for 
Purposes of Defining the Sample Population

    Several commenters also raised issues regarding the use of CBP 
data. These comments generally focused on those instances where CBP 
data may be problematic due to, for example, fraud, miscalculations, or 
multiple affiliations of sellers and resellers. Some commenters urged 
the Department to consider greater use of quantity and value (``Q&V'') 
questionnaires, while others also recognized that Q&V questionnaires 
are time-consuming and will probably lead the Department to an 
incomplete picture of the industry, especially in large industries.
    Some commenters argued that the Department should exclude producers 
with statistically insignificant export volumes (for example, less than 
two percent). Such companies' sales may not be bona fide sales, and 
selecting such companies may result in a skewed sample. These companies 
should be excluded from the sample pool while still assigning them the 
sample rate from that review. One commenter further recommended 
establishing a rebuttable presumption that entries accounting for less 
than one percent of the import volume are not bona fide sales.
    The Department's response: For the reasons explained herein, the 
Department intends to follow its current practice of relying upon CBP 
data. Consistent with that practice, the Department will consider any 
specific problems or issues identified concerning the reliability of 
CBP data on a case-by-case basis. The Department recognizes that the 
use of Q&V questionnaires is time-consuming and not always necessary 
and therefore intends to use them only where warranted, such as AD 
investigations in non-market economy countries.
    With respect to the proposal to exclude producers based on low 
export volumes, at this time, the Department does not intend to 
implement a general rule to exclude any respondents based on sales 
volumes, especially in light of utilizing the PPS methodology, which 
ensures that any single respondent is not over-represented in the 
sample population, as implementing such a singular approach would be 
inappropriate in many cases. But, the Department will consider comments 
raised by interested parties on a case-by-case basis and make 
determinations based upon the facts and circumstances in each case. The 
Department will consider all information and allegations regarding 
specific CBP data and other sales volume issues on a case-by-case 
basis.

Issue: Stratification

    Commenters questioned whether the Department should forgo 
stratification, define the strata based on different criteria than 
proposed, as well as consider defining the population (and probability 
of selection) by production, by import volume rather value, and by 
whether the respondents requested a review or whether respondents were 
named in a request for a review. One commenter argued that the 
Department has no factual basis for using size as a basis for 
stratification, which ``must be based on some relationship between the 
criteria used or the strata and the variable being measured.'' If the 
Department wishes to stratify, it must base strata on variables 
relevant to margins. One commenter proposed bifurcating the population 
into two groups: (1) Those respondents who requested a review of their 
own entries; and (2) respondents requested by the domestic parties. 
Under this novel methodology, the Department would stratify and sample 
the two populations separately, and assign rates to individual strata.
    The Department's response: The Department intends to stratify on 
the basis of volume, as this best meets the policy intentions described 
above; namely, creating the potential for individual examination for 
some of those respondents under review that otherwise would not 
normally be selected. Where circumstances warrant, especially in light 
of the enforcement concerns described herein, the Department may 
consider other characteristics by which to stratify on a case-by-case 
basis.

Issue: Whether the Department Should Limit Sampling to Reviews

    The Department also received comments regarding the use of sampling 
in investigations as well as whether sampling should be the ``default'' 
method for respondent selection. At least one commenter argued that the 
Department should use sampling as the ``default'' procedure for 
respondent selection in administrative reviews. However, given the 
complexities and short time frames of investigations, the commenter 
recommends that the Department should establish deadlines under which 
petitioners must request sampling in investigations, with ``selecting 
the largest'' as the default procedure in investigations. Other 
commenters suggest only allowing sampling in investigations when doing 
so is requested in the petition. Another group of comments recommended 
that choosing the largest should remain the Department's ``default'' 
procedure for respondent selection, given the issues to which sampling 
gives rise. Many commenters urged the Department to retain its 
discretion in choosing its respondent selection methodology as the 
facts warrant.
    The Department's position: Section 777A(c)(2)(A) of the Act 
provides the Department with authority to employ samples in both AD 
investigations and administrative reviews. The methodology described 
herein, however, was developed for purposes of administrative reviews. 
In large part, the enforcement concerns raised herein are not as 
salient in the case of investigations, where there has been no previous 
expectation of participating in (or being excluded from) a proceeding. 
Accordingly, the Department intends to consider sampling when the 
criteria described above are met in administrative reviews. Requests 
for sampling in investigations, for example, may give rise to other 
concerns that the Department has not yet considered. Therefore, the 
Department will address other requests for sampling as they arise in 
specific proceedings.

Issue: Whether the Department Should Reconsider Certain Aspects of the 
Proposed Methodology

    The Department also received comments on the methodology itself, 
with some commenters arguing that the Department should retain the 
discretion to sample when selecting only two respondents, and other 
commenters arguing that three respondents is insufficient to meet the 
statutory requirements with respect to sampling. Further, the 
Department also received

[[Page 65968]]

comments on the initially proposed 75 percent threshold, i.e., the 
percentage of imports represented by the largest respondents.
    One commenter noted that the Department should use this limitation 
(i.e., the threshold) when sampling in investigations, but not in 
reviews, since this will not address the issues sampling is intended to 
remedy in industries dominated by a few large exporters. Another 
commenter noted that the Department has not articulated any rational 
basis to reject the greater coverage of 75 percent in favor of the 
lower percentage of imports likely to be covered by a sample. Rather, 
the Department should be required to individually examine a number of 
respondents proportional to the number of respondents in the 
population.
    The Department's response: For the reasons described in greater 
detail earlier in the preamble and for purposes of this notice, the 
Department has determined to consider sampling when it can select a 
minimum of three respondents to examine individually and when the three 
largest respondents (or more if the Department intends to select more 
than three respondents) by import volume of the subject merchandise 
under review account for normally no more than 50 percent of total 
volume. The Department considers 50 percent to be a reasonable 
threshold because in these circumstances the agency would be able to 
calculate specific dumping margins for the majority of imports during a 
period of review. However, when selecting the largest respondents does 
not allow the Department to calculate dumping margins for the majority 
of imports, and the Department has the resources to review at least 
three respondents, the Department may choose to sample in view of the 
enforcement concerns discussed herein.

Issue: Respondent Characteristics

    Several commenters noted that the Department should clarify what 
information it will consider with respect to variations in the 
population. Further, while the proposed methodology does acknowledge 
that significant differences in the population may affect the decision 
to sample, it does not address how the Department will assess these 
differences. In this vein, another commenter contended that the 
comments the Department receives in the proposed 10-day deadline should 
be used by the Department not only to determine whether to sample, but 
also how to sample. Several commenters warned against relying on the 
information presented in the comments as the basis to avoid sampling.
    The Department's response: In general, the Department may consider 
sampling for respondent selection purposes in AD administrative reviews 
when (among other conditions) information obtained by or submitted to 
the Department provides a reasonable basis to believe or suspect that 
the average export prices and/or dumping margins for the largest 
exporters differ from such information that would be associated with 
the remaining exporters. Such a fact pattern supports the existence of 
potentially significant enforcement concerns, as variation in the 
dumping behavior of the population gives rise to concerns that a non-
random means of respondent selection may systematically exclude certain 
dumping behavior. The Department has identified several types of 
information that a party may submit, including: Company margins from 
previous segments of the proceeding; market and company pricing 
information; the nature and structure of the foreign industry in 
question, including cost structure and/or actual pricing data; and the 
U.S. Customs and Border Protection import entry database. The 
Department may consider other information on a case-by-case basis.

Issue: Timing

    Several commenters contended that the Department should clarify 
that the clock for the 10-day comment period should start running when 
parties have all the information necessary to submit comments (i.e., 
after the deadline for seeking separate-rate status, no-shipment 
status, Q&V/CBP data is complete, etc.). The same commenters proposed 
establishing a 40-day deadline for submitting and clarifying no-
shipment and separate-rate information, with a 10-day comment period 
following that.
    One commenter proposed waiting to sample until the window for 
withdrawing review requests has expired (currently 90 days from 
initiation), while another commenter proposed amending 19 CFR 
351.213(d)(1) to be 60 days from initiation or 15 days following the 
deadline for filing. However, these commenters also noted that the 
Department should retain discretion to adjust this deadline on a case-
by-case basis, keeping the deadline at 90 days for cases where sampling 
is not employed.
    The Department's response: The Department expects to clarify many 
of these timing issues by giving interested parties notice of the 
procedural requirements during the course of the particular proceeding, 
and will address any concerns as they arise on a case-by-case basis. In 
addition, the Department is promulgating an amendment to section 
351.301 of its regulations to address procedures for submissions 
related to sampling in administrative reviews. With respect to 
withdrawal of review requests and its potential impact on the timing of 
sampling, in cases where the Department determines to employ sampling 
for respondent selection, it will conduct its sampling following the 
conclusion of the 90-day period for withdrawal of requests for 
administrative reviews under 19 CFR 351.213(d)(1). In cases where the 
Department decides to sample, the Department expects that it will not 
exercise its discretion to extend the 90-day period for withdrawal of 
review requests. In this way, the Department preserves the ability of 
firms to withdraw their review requests during the first 90 days of the 
review as required by section 351.213(d)(1) of its regulations, but 
also ensures that later withdrawals do not adversely impact the 
Department's ability to conduct its sampling in a timely manner given 
the time constraints for completion of administrative reviews.

Issue: Rate Assignment

    One commenter maintained that the Department should assign each 
stratum's rate to the members of that stratum and should not average 
the rates together to calculate and assign a population-wide average 
rate; each stratum's rate is predictive of the behavior of members of 
that stratum, and averaging the rates together does not yield 
representative results for any member of the population.
    The Department received a range of comments regarding the inclusion 
of adverse facts available (``AFA''), de minimis and zero rates in the 
sample rate, including that: (1) The Department should include all AFA, 
zero, and de minimis margins in the sample rate; (2) the Department 
should include AFA rates and exclude de minimis/zero rates; and (3) the 
Department should exclude all total AFA, zero, and de minimis margins, 
but should include margins based on partial AFA in the sample rate.
    Several commenters submitted that the Department should use the 
weighted average of all calculated rates where there is at least one 
rate not based on AFA. Recognizing that there is no statutory directive 
when no calculated rates are available, this commenter noted that Court 
of International Trade and WTO precedent require the Department to 
``consider the

[[Page 65969]]

significance'' of zero and de minimis rates. However, these commenters 
and others further argued that international obligations are 
unambiguous with respect to this issue: AFA cannot be included in all-
other or sample rates. Article 6.8 and Annex II list limited situations 
in which AFA may be applied, and that is only when a party does not 
cooperate.
    The Department's response: As noted above, the aim of the sampling 
methodology is to obtain the population average (mean) dumping margin 
which is the trade-weighted average dumping margin across all firms 
under review. The Department considered the approaches suggested by the 
commenters, but found that the methodology described herein remains the 
most appropriate approach. The Department intends, however, to address 
any comments on how to assign rates on a case-by-case basis as they 
arise within a particular proceeding. Thus, in assigning all non-
selected companies a rate, the Department will calculate a ``sample 
rate,'' based upon an average of the rates for all selected 
respondents, weighted by the import share of their corresponding 
strata. In line with the Department's practice heretofore, the 
Department will include all rates in the sample. Therefore, consistent 
with the statute, the Department will assign one rate to all 
respondents in the sample population that were not individually 
examined. The Department will address any further issues as they relate 
to the facts of specific proceedings on a case-by-case basis.

Issue: Replacement Respondents and the Use of Voluntary Respondents

    Several commenters noted that the Department should address the 
potential need to replace a respondent. In such an event, one commenter 
suggested, the Department could rank all respondents in each stratum, 
and simply go down the list to replace a respondent. Alternatively, the 
Department can ``re-run'' the selection within that stratum. One 
commenter warned against ``re-shuffling'' the strata after a 
withdrawal, noting that the sample methodology need only be based on 
the facts known to the Department at the time of selection. Another 
commenter asserted that replacement of a respondent must be achieved 
through the PPS selection methodology in the affected stratum, 
``otherwise the sample will be skewed and any pretense of statistical 
validity will be further undermined.'' It was also noted that, if the 
Department waits to sample until the population is set (after 
withdrawals and separate-rate applications), the issue of whether to 
replace respondents should not regularly occur. One commenter stated 
that inclusion of smaller companies increases the likelihood of non-
cooperation and that the Department must increase the number of 
companies sampled in order to accommodate this eventuality. A number of 
commenters requested that the Department provide explicit guidelines 
for its selection of one or more additional mandatory respondents where 
a company initially selected does not cooperate.
    With respect to voluntary respondents, several commenters contended 
that the Department should not alter its current voluntary respondent 
practice. Further, voluntary respondents should receive their own rates 
and those rates should not be used in the weighted average rate. At 
least one commenter contended that the Department should not allow for 
voluntary respondents when sampling, but stated that if any voluntary 
respondents are examined, those rates should not be included in the 
sample rate.
    A number of commenters submitted that increasing opportunities for 
voluntary respondents provides a means to meet the Department's legal 
obligations, and that the Department's current policy of examining no 
voluntary responses whenever it has determined to limit the number of 
respondents ignores its own statute and international obligations. In 
general, these commenters urge the Department to encourage voluntary 
participation and be liberal in accepting voluntary respondents.
    The Department's response: Prior to selecting its sample, the 
Department intends to establish the population from which to draw its 
sample by first accounting for withdrawals of requests for review and 
also the separate-rate status of respondents in NME cases. However, the 
exact replacement procedure, when replacement is considered, as well as 
whether the Department will accept any specific requests for 
individual-examination by voluntary respondents, will depend, as it 
must, on the facts of the specific case. In addition, the Department 
finds the comments, such as the impact of company size on the sample, 
to be speculative at this point, but will consider such comments raised 
by interested parties in the course of its proceedings on a case-by-
case basis.

Review of the NME Entity

Background

    While considering the many issues involved in sampling in 
administrative reviews, the Department determined that one of the 
issues that may impact the use of sampling in future segments is the 
Department's review of the NME entity in its administrative reviews. 
Specifically, in proceedings involving NME countries, the Department 
has a rebuttable presumption that the export activities of all 
companies within the country are subject to government control and, 
thus, imports from all companies should be assessed a single AD rate 
(i.e., the NME-entity rate).\18\ It is the Department's practice to 
assign this single rate to all exporters of merchandise in an NME 
country subject to an AD investigation or review unless an exporter can 
demonstrate that it is sufficiently independent in its export 
activities, on both a de jure and de facto basis, so as to be entitled 
to a ``separate rate'' (i.e., a dumping margin separate from the margin 
assigned to the NME entity). The Department analyzes each entity 
exporting the subject merchandise that applies for a separate rate 
under a test first articulated in Sparklers,\19\ and further developed 
in Silicon Carbide.\20\
---------------------------------------------------------------------------

    \18\ See 19 CFR 351.107(d) (providing that ``in an antidumping 
proceeding involving imports from a nonmarket economy country, 
`rates' may consist of a single dumping margin applicable to all 
exporters and producers'').
    \19\ See Final Determination of Sales at Less Than Fair Value: 
Sparklers from the People's Republic of China, 56 FR 20588 (May 6, 
1991) (``Sparklers'').
    \20\ See Final Determination of Sales at Less Than Fair Value: 
Silicon Carbide from the People's Republic of China, 59 FR 22585 
(May 2, 1994) (``Silicon Carbide'').
---------------------------------------------------------------------------

    Exporters named in the initiation of an AD administrative review 
that do not establish that they are independent of government control 
are considered part of the NME entity. In such instances, it has been 
the Department's practice to consider the NME entity under review, even 
if no request for review was made specifically for the entity.\21\ 
Under this practice, the assessment rate for entries from exporters 
that are part of the NME entity is not determined until the final 
results of the review. Thus, the Department typically does not instruct 
CBP to liquidate entries for any exporters whose deposits were made at 
the rate of the NME entity pending the final results of the 
administrative review. As a result, importers with entries from 
exporters that are part of the NME entity, but that were not named in 
the initiation of the review,

[[Page 65970]]

must nevertheless wait until the final results of review before final 
liquidation. However, in most cases, the assessment rate is not 
different from the cash deposit rate at the time of entry for such 
imports. Consequently, the Department's conditional review practice has 
resulted in the delayed liquidation (often over a year after the date 
of initiation) of NME entity entries, even though the NME entity rate 
is unlikely to change when the NME entity is under review.
---------------------------------------------------------------------------

    \21\ This practice was affirmed in Transcom, Inc., v. United 
States, 294 F.3d 1371 (Fed. Cir. 2002).
---------------------------------------------------------------------------

Statement of Practice Regarding Review of the NME Entity

    The Department will no longer consider the NME entity as an 
exporter conditionally subject to administrative reviews. Accordingly, 
the NME entity will not be under review unless the Department 
specifically receives a request for, or self-initiates, a review of the 
NME entity.\22\ In administrative reviews of AD orders from NME 
countries where a review of the NME entity has not been initiated, but 
where an individual exporter for which a review was initiated does not 
qualify for a separate rate, the Department will issue a final decision 
indicating that the company in question is part of the NME entity. 
However, in that situation, because no review of the NME entity was 
conducted, the NME entity's entries were not subject to the review and 
the rate for the-NME entity is not subject to change as a result of 
that review (although the rate for the individual exporter may change 
as a function of the finding that the exporter is part of the NME 
entity).
---------------------------------------------------------------------------

    \22\ In accordance with 19 CFR 351.213(b)(1), parties should 
specify that they are requesting a review of entries from exporters 
comprising the entity, and to the extent possible, include the names 
of such exporters in their request.
---------------------------------------------------------------------------

    Following initiation of an administrative review when there is no 
review requested of the NME entity, the Department will instruct CBP to 
liquidate entries for all exporters not named in the initiation notice, 
including those that were suspended at the NME entity rate. This change 
in practice will eliminate the unnecessary delay in liquidation of 
entries from the NME entity.

    Dated: September 30, 2013.
Paul Piquado,
Assistant Secretary for Import Administration.
[FR Doc. 2013-26266 Filed 11-1-13; 8:45 am]
BILLING CODE 3510-DS-P